On Tue, Oct 3, 2017 at 01:41 pm, Alan G4ZFQ wrote:
Without actual power figures it is difficult to tell.I found some notes I took some time back while burning up power transistors attempting the "peak voltage minus some" biasing method with the dummy load, which I do not recommend and especially not without proper heatsinking of the BS170 transistor.
Considering I was trying to get the figures written down and back off the bias before the transistor popped, these are somewhat disorganized, but I see that the peak voltage on 40m was 5.4VDC [!] and the peak voltage on the other U3S running 20m was 3.9VDC when these U3Ss were transmitting (the meter used is my best and believed to be very accurate from other testing and comparisons).
By Hans's calculations with these devices into the identical dummy load, 5.4VDC translates to approximately 425mW and 3.9VDC to about 275mW, both very higher than expected. I am not sure, but I believe these were both using the very much stronger Fairchild BS170s I got, which in all my testing seem to produce about 20%-25% greater output than whatever it is that comes in the kits in the identical situations. One U3S had a monofilar coil and one had the bifilar coil, but I couldn't swear to which was which. 40m load biasing cooked transistors so fast it wasn't even funny.
I did not test with 30m and 17m frequencies and LPFs.
I suspect these tests would bear more organized and complete repetition, perhaps with my new oscilloscope that's been staring at me for the past six weeks, patiently awaiting its initial use.